tocmo0nlord/axolotl
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

replace references to random 68m model w 135m smollm2 (#2570) [skip ci]

Browse Source 
* replace references to random 68m model w 135m smollm2

* use AutoTokenizer for smollm2
This commit is contained in:
Wing Lian
2025-04-28 10:08:07 -04:00
committed by GitHub
parent f1df73a798
commit 40f4ea23ab
13 changed files with 95 additions and 87 deletions
Download Patch File Download Diff File Expand all files Collapse all files

28
tests/e2e/test_optimizers.py
View File

@@ -28,8 +28,9 @@ class TestCustomOptimizers(unittest.TestCase):
# pylint: disable=duplicate-code
cfg = DictDefault(
{
"base_model": "JackFram/llama-68m",
"tokenizer_type": "LlamaTokenizer",
"base_model": "HuggingFaceTB/SmolLM2-135M",
"model_type": "AutoModelForCausalLM",
"tokenizer_type": "AutoTokenizer",
"sequence_len": 1024,
"load_in_8bit": True,
"adapter": "lora",
@@ -39,9 +40,7 @@ class TestCustomOptimizers(unittest.TestCase):
"lora_target_linear": True,
"val_set_size": 0.02,
"special_tokens": {
"unk_token": "<unk>",
"bos_token": "<s>",
"eos_token": "</s>",
"pad_token": "<|endoftext|>",
},
"datasets": [
{
@@ -75,8 +74,9 @@ class TestCustomOptimizers(unittest.TestCase):
# pylint: disable=duplicate-code
cfg = DictDefault(
{
"base_model": "JackFram/llama-68m",
"tokenizer_type": "LlamaTokenizer",
"base_model": "HuggingFaceTB/SmolLM2-135M",
"model_type": "AutoModelForCausalLM",
"tokenizer_type": "AutoTokenizer",
"sequence_len": 1024,
"load_in_8bit": True,
"adapter": "lora",
@@ -86,9 +86,7 @@ class TestCustomOptimizers(unittest.TestCase):
"lora_target_linear": True,
"val_set_size": 0.02,
"special_tokens": {
"unk_token": "<unk>",
"bos_token": "<s>",
"eos_token": "</s>",
"pad_token": "<|endoftext|>",
},
"datasets": [
{
@@ -122,8 +120,9 @@ class TestCustomOptimizers(unittest.TestCase):
# pylint: disable=duplicate-code
cfg = DictDefault(
{
"base_model": "JackFram/llama-68m",
"tokenizer_type": "LlamaTokenizer",
"base_model": "HuggingFaceTB/SmolLM2-135M",
"model_type": "AutoModelForCausalLM",
"tokenizer_type": "AutoTokenizer",
"sequence_len": 1024,
"load_in_8bit": True,
"adapter": "lora",
@@ -133,9 +132,7 @@ class TestCustomOptimizers(unittest.TestCase):
"lora_target_linear": True,
"val_set_size": 0.02,
"special_tokens": {
"unk_token": "<unk>",
"bos_token": "<s>",
"eos_token": "</s>",
"pad_token": "<|endoftext|>",
},
"datasets": [
{
@@ -170,6 +167,7 @@ class TestCustomOptimizers(unittest.TestCase):
cfg = DictDefault(
{
"base_model": "HuggingFaceTB/SmolLM2-135M",
"model_type": "AutoModelForCausalLM",
"sequence_len": 1024,
"val_set_size": 0.01,
"special_tokens": {